HematoCARD: A Volumetric Duplicate Dried Blood Spot Collection Cartridge Validated for Therapeutic Drug Monitoring

The growing global demand for in vitro diagnostic testing requires the resource-intensive, centralized healthcare system to convert to a more decentralized, patient-centric model. In this context, patient-centric sampling through dried blood spot (DBS) collection is emerging as one of the solutions to facilitate this shift. This method is intended for the collection of capillary blood through finger-pricking and has the potential to reduce invasiveness and improve user-friendliness while enabling individuals to perform self-sampling in low-resource settings (e.g., at home). However, DBS sampling still suffers from significant inaccuracies primarily caused by variations in blood hematocrit levels, spot-to-spot variability, and user-dependent errors such as under- or overloading, wrong positioning, or contamination. To address these issues, this work introduces a user-friendly, self-powered microfluidic-based cartridge, named HematoCARD. The cartridge autonomously collects metered duplicate DBS samples of precisely 10 μL each from a single finger-prick sample. Additionally, it integrates a "smart" inlet system designed to be user-friendly and to avoid user-dependent errors. The HematoCARD was validated for monitoring the therapeutic antibody adalimumab by testing spiked blood samples, demonstrating good analytical performance in terms of accuracy (91-111%), linearity (R² > 0.99), and variability (CV < 15%). The device was shown to outperform commercial porous-based microsampling systems in terms of accuracy and hematocrit dependency while matching with microfluidic-based systems. Clinical validation showed good correlation (Pearson's r ≥ 0.94) with serum reference and commercial DBS microsampling methods while providing a clear advantage of reduced spot-to-spot variability through replicate sampling. These advancements are an important step forward in patient-centric sampling technology, offering improved parallelization, increased robustness, and enhanced practical usability of self-sampling devices.